Towards shorter outages in Carolina1 May 2000
Staff mobility, new technologies and football analogies are reducing outages to under 30 days at Carolina Power & Light.
Ten years ago, it was usual for refuelling and maintenance outages at Carolina Power & Light’s (CP&L’s) four nuclear units to span two months or more, as it was across the whole of the US. A common sight at CP&L plants during these protracted outages was a huge number of contractors on site performing, in many cases, turnkey operations. Long outages and high staffing levels combined to leave CP&L’s nuclear programme with low capacity factors and high operating and maintenance costs.
Since then, the company has worked to reverse those trends, with dramatic results. In 1999, CP&L's Brunswick 2 (GE BWR, 811MWe) at Southport, North Carolina, and Robinson (Westinghouse PWR, 683MWe) at Hartsville, South Carolina, each set unit records for outage efficiency. Brunswick 2 completed its 14th outage in 36 days during the spring, and in the fall, Robinson set a company standard, closing out its 19th refuelling outage in just 29 days.
Those scheduled outages, records for the company, and a maximum of on-line time at CP&L's nuclear plants (despite two hurricane-forced shutdowns at Brunswick), combined to produce a net generation in excess of 26TWh. This was CP&L’s sixth consecutive annual record for nuclear generation. And even with two refuelling outages, the company’s nuclear group capacity factor for 1999 was 93.6%.
Robinson’s outage record broke the previous site mark by 11 days and continued a downward trend established throughout the 1990s. At 29 days, refuelling outage 19 took less than half the average time (73 days) of all outages since the plant began commercial operation in 1971. (The average does not include replacement of three Westinghouse steam generators – 350 days – during the plant’s ninth outage in 1984.)
The company now plans for each of its nuclear units (in the absence of major plant modifications) to keep outages to below 30 days. That goal includes rigorously enforced deadlines for individual work items and planning milestone schedules for future outages. In many cases, targeted work-item schedules are considerably shorter than the existing unit record.
CP&L will not be able to keep all future outages shorter than 30 days. Plant maintenance conditions, related improvement initiatives and the plant’s life-cycle management, in large part, will continue to dictate prospective and actual outage duration.
Plans for Brunswick 1 (GE BWR, 820MWe) which, like its sister, has gone from an 18-month refuelling cycle to a 24-month cycle, called for a 30-day outage early this year (the unit’s 13th outage), and Harris (Westinghouse PWR, 860MWe) at New Hill, North Carolina, is planning its ninth refuelling outage for 28 days.
Among the efforts that have contributed to the decrease in outage duration is the sharing of human resources among all CP&L plants. Experience has shown that one shared resource employee can typically replace the need for two contractors, since the CP&L employee is already qualified, requires less training and is usually more familiar with CP&L work processes. By standardising many practices and procedures, CP&L has increased the mobility of employees between sites.This has led to a dramatic decrease in the numbers of contractors needed during outages.
Sharing resources among plants provides numerous benefits. Since the workers are CP&L employees, they receive bonuses for short, error-free outages. Additionally, as the workers repeat their outage tasks, they become better at them, resulting in additional time savings. By using expert teams, CP&L has reduced contractor numbers at the sites and improved efficiency significantly.
In the recent Brunswick 2 outage, for example, a reactor services crew comprising 20% fewer contractors than ever before completed the reactor vessel disassembly 12 hours ahead of schedule with no human performance issues.
Robinson's early plant design can hamper the speed of certain refuelling processes, so plant employees have worked to identify and develop process efficiencies, and at times, upgrade equipment and systems that are key to shortening outage times. With a premium on getting maintenance and refuelling right the first time, detailed advance planning and scheduling are critical.
When developing a schedule, planners use historical outage data on each major activity to develop a backbone schedule based on the best times to complete that activity. Employee feedback, generated through post-outage debriefings, helps determine opportunities for improvement. And a significantly higher portion of what were traditional outage tasks are now performed with the plant on line. This has resulted in significant time savings and has served to increase employee focus on the most critical outage tasks.
Among the technological components of Robinson’s record outage was the use of a new type of cavity seal. The old style was inflatable and maintained by air or nitrogen, which threatened a risk of failure when the refuelling cavity was filled with water. The old seal also took 18 to 24 hours of critical-path time to install and remove, and due to the length of time they spent next to the reactor, radiation doses to employees were significant.
The new seal came in four major pieces and was installed in less than an hour. The rubber T-type seal locks down to hold the seal in place with little or no probability of failure. Water leakage from the cavity was the lowest in the plant’s history.
To reduce the outage time required to vent and drain steam generator tubes, so that work could be performed in the loops and on the reactor coolant pumps, nitrogen was injected into the loops to displace the water in the tubes. While this method had been used elsewhere in the industry, this was its first use at Robinson.
The process resulted in a saving of about 18 hours, a saving that will be realised in future outages as well.
Employee commitment has also improved. Operations personnel, for instance, found ways to increase the efficiency of processes and methods used to shut down and cool down the plant.
Project teams had been assigned a year in advance to allow employees to prepare for the major outage projects. The assignments resulted in numerous outage improvements, including enhanced ownership and better handoffs. A new software program facilitated handoffs among work groups, allowing any employee to look at the overall schedule of outage tasks, to better understand the chronology and facilitate efficient project movement.
Maintenance management assigned work to individual teams of two to three people. These personnel reviewed work tickets, knew where they were supposed to work, and prepared their tools and supplies so that they could begin that activity as soon as the schedule and other work allowed it.
Days before the outage was to start, the threat of a direct hit by Hurricane Floyd forced Robinson employees to remove much of the equipment and materials that had been pre-staged for the work. Then, on the fourth night of the outage, a tornado touched down on site. The twister damaged 50 site vehicles, snapped trees and damaged a temporary canopy that covered the turbine deck. But employee teams responded quickly, and the storm did not have any effect on the outage schedule.
Brunswick personnel have reduced the outage work scope by developing a maintenance rule that allows them to manage each system’s out-of-service time, while allowing scheduled online outages for components. Following completion of safety analyses, maintenance management has moved several maintenance surveillance tests (MSTs) from a refuelling outage schedule to an online schedule.
Like Robinson, Brunswick employees use the company’s own scheduling and status tool, Progress, to provide real-time status of outage activities, in an effort to manage resources efficiently. The Progress tool shows such activities as scaffolding installation and removal, post-maintenance testing and so on in detail.
CP&L has received inquiries from a number of other nuclear utilities, seeking to purchase the Progress system.
Plant personnel now scrutinise each job repeatedly during the planning process, to ensure that it is performed at the optimal point in the schedule. Emergency bus work, for instance, has been sequenced in with similar work, rather than scheduled at the end of the outage as critical path, as had been the standard practice.
Contingency planning, a much larger part of outage planning and scheduling than ever before, has already paid dividends.
During Brunswick 2’s outage last year, there was a main generator busbar test failure that had an original repair estimate of 15 days. With a contingency plan in place to accommodate such a problem, however, the repair was completed in eight days, allowing Brunswick to meet its business plan schedule for the outage.
Staff at Brunswick are now empowered to make decisions as issues arise, rather than awaiting management. There are four rotating outage teams at the plant, each comprising representatives from maintenance, engineering, radiation control and shift outage management.
Communication among work groups and between management and employees, as well as recognition at all levels of the organisation, is increasingly important. Given the increased demands on employee time and the heightened focus on efficiency, before and during outages, the company emphasises recognising and rewarding employees for resourcefulness, commitment and innovation.
Outages at CP&L’s plants use strategic themes as a means of focusing employee effort and rallying support for the common cause of a safe, efficient return to Mode 1. During Robinson’s recent outage, for instance, an American football theme focused on the individual and collective performance mastered by successful sports teams.
On a master gridiron tote board, “first-downs” were awarded for such milestones as RCS cooldown and depressurisation, completion of fuel offload and core reload. A touchdown was awarded when the unit achieved 100% power. The motif is intended to reinforce the concepts, practices and power of teamwork and ownership among employees.
More pragmatically, communication has been improved through the addition of radios on site. At Brunswick, all supervisors, outage team members and coordinators have dedicated radios to allow instant communication with fellow team members. Cellular telephones are now widely used on site at Robinson.
While planning and process changes have reduced 60-day outages to outages in the 30-day range, incremental improvements beyond the 30-day mark require year-round focus. Earlier this year, CP&L embarked on a programme aimed at achieving further improvements by bringing about a cultural change within the nuclear operations and support organisations. Titled ‘Zero Tolerance 2000,’ the programme focuses on a continuously improving schedule of maintenance.
Zero Tolerance 2000 focuses on preventing equipment failures from happening in the first place. Patterned after cultures witnessed at a number of European nuclear plants, the programme aims to identify and understand the life cycle of integral plant systems and equipment, and to repair, replace or rebuild equipment before it fails. It was rolled out early this year by chief nuclear officer CS "Scotty" Hinnant.
In the rollout at each CP&L nuclear site, Hinnant quoted a report on the European predictive maintenance culture from INPO 99-002. He observed an extreme intolerance for equipment problems at each of the stations. Corrective maintenance activities were considered ‘failures’ and too expensive to be tolerated. Personnel initiated cause analysis for repetitive corrective maintenance or tasks that have untimely resolution, and addressed most corrective maintenance in days.
Each of the stations had both a low failure rate and a corrective maintenance backlog ranging from 20 to 50 items.
The CP&L plan calls for identifying a master list of equipment critical to, or required for, 100% power operation. Maintenance attention will focus on that equipment. Other less-critical equipment will be designated as run-to-failure and will not be routinely monitored by a predictive or periodic maintenance programme.
For equipment on the master list, CP&L is developing a detailed understanding of its operating history, life expectancy and critical characteristics. The company will then implement a maintenance strategy, including time-based rebuilds, to improve equipment reliability and availability.
Preventive maintenance, including predictive (condition-based) and periodic (time-based) activities, focuses on maintaining equipment within design operating conditions to extend operating life.
The predictive maintenance programme calls for the use of numerous key technologies, including vibration monitoring and diagnostics, lubricating oil and grease analysis, infrared thermography imaging, and operating equipment observation.
Such cultural changes – as part of an arsenal of planning and maintenance activities – will become increasingly important as Southeast electricity markets deregulate. As with other parts of the US and Europe, the regulatory and business landscapes are constantly changing.
Later this year, CP&L anticipates completing its acquisition of Florida Progress, parent company of Florida Power, which owns Crystal River (Babcock & Wilcox PWR, 885MWe).
The acquisition offers logistical challenges, not least Crystal River 3’s distance from the Carolinas, but promises opportunities as well, with expanded shared resource capability and heightened exchange of best practices. Crystal River 3 is already on the way to improved outage performance, establishing a unit record in the fall of 1999 with a 42-day outage, during which workers made more than 60 plant modifications.